UNFPAState of World Population 2002
Back to Main Menu

Chapter 4:

Sources of City Growth

The tendency towards city growth, both in absolute terms and as a proportion of populations, is influenced by differing patterns of fertility, mortality and migration.

Natural Increase

The first cause of urban population growth is simply an excess of births over deaths, known as natural increase. Urban fertility is usually lower than rural fertility though the differences vary from region to region. Asia shows smaller differences than Africa and Latin America. In Asia, the difference ranges from about a half to one birth over the reproductive span, in sub-Saharan Africa from one to two; in Western Asia and North Africa, around two; in Latin America between one and three. These differences have emerged relatively recently, particularly in sub-Saharan Africa.1

High population growth rates in the rural areas help fuel migration: many of the migrants are in the prime of their reproductive years and their children are added to the city populations. It is a question, however, whether the fertility of migrants follows the pattern of their origin or that of their destination.

The available data suggest that migrants quite quickly adopt their new neighbours' fertility behaviour: urban natives have the lowest fertility, then rural to urban migrants, then urban to rural migrants, with rural natives the highest of all. These differences are more marked than they were a decade ago when urban areas were less advanced in the process of demographic transition.

Differences between regions are the result of circumstances as well as demographic change. Urban migrants in Africa are often separated from their wives or husbands: their fertility is extremely low in the first two years and then increases until the overall rate is about the urban norm after five years. In Latin America the same initial pattern stabilized over time at lower levels; it remains to be seen if this will happen in Africa. Recent analysis suggests that it will, because of better employment and other opportunities and more accessible health services in urban areas.2

Women in cities generally tend to marry later, breast-feed less and abstain from sex for a shorter time after a birth; but they also have better access to reproductive health and family planning services and are more likely to use contraception. But within this generalization there are wide variations among individuals and groups, depending on cultural background, economic and social opportunity, education levels, service access and aspirations.

Rural and urban areas differ not only in fertility but in mortality. Overall, urban areas tend to have lower levels of mortality than rural areas, though mortality in poorer urban neighbourhoods can be at least as high as in rural areas. Even in developed countries, high levels of infant and child mortality can be observed in poor urban neighbourhoods. In combination with high levels of adult mortality due to violence and accidents, life expectancies can be extremely low. The mortality rate in the Harlem section of New York from birth to age 65 is estimated to be higher than the national average for Bangladesh. Overall comparisons of rates of natural increase in urban and rural areas are not readily available. The most systematic attempt to perform the necessary analyses is now over 15 years old.3

Over the last three decades of declining mortality and fertility, the relative importance of natural increase has remained at about 60 per cent of urban growth. In the 1960s, this was the level in all major regions of the world. By the 1980s, however, the dynamics of development led to significant regional differences. In Asia, migration and reclassification started to become more important in the 1970s and by the 1980s had reached nearly half.4 During the difficult decade of the 1980s in Latin America, natural increase accounted for two thirds of urban growth, and in Africa for three quarters.

These regional differences reflect trends in component processes. In Latin America, urban natural increase has been declining while migration has fluctuated (higher in the more successful decade of the 1970s than before or since). In Asia, migration has steadily increased since the 1960s while natural increase grew slightly over time.5 In Africa, natural increase has remained high, but migration rates dropped by half.

It is hard to draw country-specific policy conclusions from the data available. Generally speaking, since high rural fertility rates help fuel rural to urban migration, reducing unwanted fertility in rural areas will help reduce rates of urban growth. But since natural increase is still the major component of rapid urban growth, information and services are also needed to reduce unwanted urban fertility.

Particular attention should be given to peri-urban and squatter settlements where new migrants tend to be concentrated.

Access to quality reproductive health services, including family planning methods, and the unmet demand for such services vary considerably between regions. Outside sub-Saharan Africa, unmet demand for family planning services tends to be higher in rural areas than in cities.6 In Latin America and the Caribbean, the urban unmet need is estimated to be low, generally under 15 per cent; in rural areas, unmet need is at least 60 per cent higher. In Asia and North Africa, rural unmet need is generally higher than in cities, but the differential is much smaller than in Latin America. Urban unmet need is as high as 32.5 per cent in Pakistan, 23.4 per cent in the Philippines and 19.9 per cent in Jordan. In sub-Saharan Africa, however, more than half the surveyed countries show higher unmet need in cities than in rural areas. Since contraceptive prevalence is higher in urban areas, this indicates that urban demand has risen and that programmes still must overcome various obstacles – financial, informational and social – to meet this increased demand.


Migration is difficult to study, since a complete analysis requires information on people and conditions in at least two different places and at various times. Information about communities, families and individuals is needed to understand the context of migratory moves. Many people move seasonally (farmers seeking non-agricultural jobs during the dry season, for example) or temporarily (to gain an education or save a certain amount of money); others make multiple moves in hopes of improving their lives.

Temporary migration is difficult to quantify without specialized surveys or studies, of which there are few; census estimates provide only snapshots of a more dynamic flow of people. Much research and policy attention is focused on rural to urban migration. In parts of the world such as Latin America, however, rural to rural migration is a much larger flow. Migration between cities and from cities back to the countryside further complicate efforts to understand migration based on information on peoples' locations at census times.

Where specialized studies have been undertaken, they have found a much greater volume of migration than census estimates suggest. A survey in Thailand, for example, found that about one person in four migrated within the previous five years, 15 per cent within the previous two years. About one third of all internal migration in the country was temporary. Other studies suggest that Bangkok's population varies by as much as 1 million people in different seasons.7

Migration can create both problems and benefits for both the areas that people migrate from and the areas they migrate to. Developing policies that can reduce the problems and maximize the benefits will require better understanding of its causes and consequences.

Why people move

Migration makes a significant contribution to the growth of urban areas. The decision for a family member or an entire family to move is a complex one, affected by both the "push" of conditions at home and the "pull" of life in the city. The mix varies from time to time and place to place; and the influence of different factors also varies according to family size and composition. The discussion here, however, concentrates on voluntary migration. In parts of the world affected by political instability, natural disasters or wars, large numbers of people become refugees within their own countries, frequently moving to cities from unsafe countrysides. In such cases "push" factors predominate as the motivation to move, though the "pull" of the cities may determine the destination.

The hope or expectation of jobs. The primary reason for urban migration has long been understood as the desire for higher wages or the greater expectation of employment. However, the difference between urban and rural wages is minimal in some settings (for example in sub-Saharan Africa). It is also widely recognized that the informal sector now accounts for a substantial portion of employment and employment growth in urban areas. The prospects of work and wages in the informal sector are much less easy to define, especially at a distance, so it is hard to account for the continuing inflow in this manner. However, better and more varied opportunities for the better-educated are reflected in higher migration rates among this group, particularly as regards men.

Low barriers to migration. Physical barriers such as distance and accessibility are reduced by better roads and transport; but lower barriers equally may encourage seasonal or circular migration. Social barriers are reduced by networks of friends or relatives who provide a context of familiarity, encouragement and support for the would-be migrant. Direct and indirect controls such as taxes, rationing, pass systems and policing slow migration, although they are frequently evaded when motivation is strong. Conversely, freedom of movement can increase migration, although movement is more likely to be temporary if it is free.

Services and amenities are generally better in the city. Educational opportunities – particularly at levels above basic education – and health services are more accessible and of higher quality. Urban life offers the prospect of freedom from gathering wood and carrying water, though many migrants must be sorely disappointed. Life in the favelas, barrios and shanty towns can be every bit as arduous and services just as poor as in the village. The quality of water and air and the risk of disease from inadequate sanitation and overcrowding may even be worse. What is known about conditions in such settlements can hardly be an inducement.

Deterioration of rural life. Rapid population growth among low-income groups in rural areas puts pressure on land, fuel and water. These pressures may be intensified by large-scale resource-intensive agriculture, loss of traditional tenure rights on common land and other policies which reduce the need for labour and the possibility of self- sufficiency. Such developments may loosen social cohesion and the sense of community as they change economic relationships. This together with pollution and degradation of basic resources reduces the quality of rural life and forces migration to the cities.

The influence of the different factors is unknown, but the erosion of traditional "pull" factors and the intensification of the "push" out of rural areas points to a generally negative conclusion in many areas, particularly sub-Saharan Africa. The balance may be different in much of Asia, where economic opportunities are generally better.

The gender factor

Studies of migration from the 1960s to the 1980s generally ignored gender issues, but migration decisions are more complex than the individual decision of a male employment seeker. Newer studies show that a growing proportion of rural to urban migration streams is made up of women and that individuals who migrate, men and women, are often doing so as part of a complex family and community process aimed at improving family well-being and survival.8 Women migrants are particularly disadvantaged by a number of recent developments which the new studies have revealed.

Migration is highly concentrated around the time of entry into the labour force, between the ages of 15 and 24.9 Recently, a trend among women towards migration at younger ages has developed, concentrated between 10 and 20. There is also a secondary peak among women in their late 50s and 60s, who commonly migrate more frequently than men in the same age group. They are often widows and divorcees moving to join their children, or childless women. The absence of sons or a husband can deny an older woman status in her own community, forcing her to seek the protection of other family members or to strike out on her own.

Studies in Asia and in sub-Saharan Africa suggest that the availability of low-skill domestic work in urban areas increasingly draws younger women. Studies in Latin America and Asia have found young women migrants supporting rural families, saving for future marriage or looking for a husband. Among poor women who migrate, many are currently single with a small number of children to support. Some of these young women return to start married life: however, it is widely believed that permanent migration is increasing for both men and women, together or apart.

Many women migrants report that they moved to accompany or follow a spouse or other family member:10 globally, about 29 per cent (16 per cent to accompany family and about 13 per cent for marriage); about a third in Asia, higher in South Asia and Africa. Such migration predominates in largely Moslem societies where women are usually married and do not join the labour force. Elsewhere, job and education, support of children and occupational mobility account for the majority. In Latin America and the Caribbean and in the South Pacific, moves for employment are the most common. Changes in economic opportunities in agriculture in rural areas have increased rural–rural migration among women in South Asia and are increasing the proportion of women's overall mobility that is believed to be economically motivated. Whether such change will also be reflected in rural–urban migration remains to be seen, but it is a reasonable response to economic conditions.

In South-east Asia and Latin America, a high proportion of women migrants surveyed cite economic motivations (68 per cent in Bangkok, 70 per cent in Dugupan City, the Philippines, 56 per cent in Malaysia, over 50 per cent in Costa Rica, Guatemala, Haiti and Mexico). The high levels of economically motivated migration in South-east Asia are relatively recent. They are related to many changes, including girls staying longer at school and better opportunities for women's employment in urban areas, especially in service sectors. The greater importance of economic motivation may also be the result of increasing independence in choosing a mate. Women in many parts of Asia are getting married later in life, so moving in order to accompany a husband or family is less likely.

Some observers have argued that the relatively wealthy and relatively poor are disproportionately represented, particularly among male migrants, but the evidence is far from clear. Women migrants, however, are more likely to be from poorer (and larger) families. Poor rural families are more likely to allow or demand female migration to cities as part of broad survival strategies. The transformation of rural economies has created a demand for cash, but not the jobs to supply it. In this case, the wages of a daughter in an office, shop or domestic job can fill the need. Often the opportunities are easier for women to find than they are for men. Even where this is not the case, experience has shown that daughters are more likely than sons to send money home.

Female migration decisions are the result of a more complex mix of social and economic conditions than male decisions. Women migrants tend to have less education than their male counterparts but this reflects the general difference in women's opportunities, and they have more education than women who do not migrate, even in societies with low levels of female education. Women with more education are more likely to find formal sector employment, and so are favoured in family migration decisions. Nevertheless these women are at a disadvantage when they reach the city, and may find themselves trapped in low-wage employment without the qualifications to advance, but still with too much education for work at home.

For most women there is little change in the type of work they do after they migrate. A third to a half of all female urban migrants find jobs in domestic or personal service. The next most common occupation is sales, either as an independent operator or employed by others. Fewer than one fourth find jobs in blue or white collar formal jobs. Trading unpaid domestic labour at home for the same work (though paid) in the city, or cultivating vegetables in order to sell them does nothing real for the autonomy of these women. The informal sector offers little security and restricted opportunities. Work such as prostitution exploits women with limited education and impoverished backgrounds. The exploitation of children – both girls and boys – for their labour is also a problem in cities, among both migrants and other residents. And in poorer families some children are abandoned to their own devices or flee difficult conditions and become "street children", cut off from family support networks and opportunities for education.

Opportunities for advancement depend on the skills which women bring with them to their urban milieu and their resourcefulness in acquiring new skills after they arrive. Though some women are aware of the problem and manage to avoid it by seeking further education once they arrive in the city, the improvement of both education and employment opportunities for women in rural areas is essential.

Unanswered questions

A broad range of questions about the contribution of migration to urban growth, the dynamics of migration and prospects for the future remain unanswered because of the difficulties of research and the sketchiness of existing data. Conceptual problems and measurement difficulties obscure conclusions on many fundamental issues. For example, the proportion of rural to urban migration which is temporary or targeted to particular short-term ends (such as seeking a marriage-partner or start-up funds) and the scale and impact of return migration are largely unknown. The diversity of factors (within families, within sending areas and within urban areas) which influence migration decisions make generalization difficult. Migration has contributed significantly to the growth of cities but an analytic framework for understanding its dynamics, including its gender aspects, remains elusive.

We know that large numbers of migrants, particularly to the larger cities, facilitate further migration. Groups of migrants once settled ease the transition of their families, associates and neighbours and reduce barriers to their movement. Studies show that a substantial majority of rural to urban migrants in most parts of the world find employment and housing through such connections. Some Asian studies found that some three quarters of migrants had employment leads before they moved.11 Such networks strengthen links with the rural places of origin, and may assist the transformation to a cash economy and the adoption of urban values. As this proceeds the migration calculus will become more complex and its effects further-reaching.

City dynamics and international migration

International migration also fuels urban growth. In places with low rates of natural increase and modest migration flows, international migration can be a significant contributor to urban growth. Population flows from rural to urban areas reflect demand for labour in urban areas, low rural opportunity and unbalanced patterns of development between rural and urban areas; similarly, international migration flows reflect broader patterns of international imbalances in development. The barriers to international migration are higher – greater expense, longer distances (commonly), greater separation from familiar surrounds, looser ties to familiar persons and practices, varying levels of legal, social and cultural acceptance by the hosts. These barriers vary in intensity, as do the formal and informal mechanisms which can reduce them.

Cities are frequently the starting and ending points of long-distance migration (particularly if it is inter-continental), though international migrants often bypass cities in their own countries. Migration flows within countries far exceed those between nations. This is probably clearest both in countries at the highest and the lowest overall development levels.12 But many of the same questions remain to be answered. Which international migrants intend permanent settlement and which intend to return to their own countries? How would that change if controls were stiffer or more relaxed? For example, some North African migrants would prefer to come and go between France or Italy and their own countries. Will migration controls keep them out or lock them in? Will older migrants retire in their countries of origin, as many "new Australians" have done?


The growth of cities is part of a secular shift in societies and economies, on a scale never before experienced. While the quantitative outline of the future is clear in general terms, the pace and intensity of the evolution and the resulting quality of life and opportunity is not. Responses are required from all sectors of civil society, to meet existing needs and to anticipate new ones, to help to create a cohesive society. Among them will be special attention to the needs and the contribution of women.

The preceding chapters have emphasized the differences between rural and urban areas. It is clear, however, that urban issues cannot be addressed in isolation, because the links between city and country are closer and more comprehensive than ever. The section which follows reviews the growing similarities as both rural and urban areas increase their linkages, respond to each other and to broader trends of an increasingly international economic environment. The final section will call attention to some of the policy responses which will be needed in an increasingly urbanized world.